while two others, Nos. 54 and 873, who were orig- 
inal members, proved to have only marginal asso- 
ciation. This analysis corroborates other lines of 
evidence, that group size among rats rarely exceeds 
10 or 12 individuals. 
The 10 rats with the greatest accumulated total 
of associations in Columns A to C were considered 
as the real members of this group, and a tabulation 
of the associations of the population with them was 
prepared. Forty-five rats were never observed to 
have any associations with these 10 rats; 100 had 
an average of 3.9 paired associations each; while the 
10 members of the group had an average of 11.5 
associations with the other members. 
Organization of the population into relatively 
discrete groups increases the number of associations 
that the members of a group have with each other. 
It also increases the number of rats with whom the 
members of the group have no association. 
E. Contacts between Groups. Rats do organize into 
fairly discrete groups with stability through time. 
However, any individual may transfer membership 
from one group to another, or it may simultaneously 
have varying degrees of association with, or affinity 
to, two or more groups. When any individual has 
contacts with members of two groups there exists 
the possibility of transfer of information between 
the two groups. Such individuals may be desig- 
nated as potential messengers. Information is here 
broadly defined as any state or condition charac- 
terizing one or more rats. Disease organisms, or 
behavioral traits are such types of information, 
which appeared to be transferred from one indi- 
vidual to another. This raises the question: How 
does the organization of the population into in- 
groups alter the possible effectiveness of messengers? 
In order to approximate an answer to this ques- 
tion, a further study was made of the four suspected 
ingroups, and the four groups whose members were 
selected at random. This study was of the number 
of contacts or paired associations which each mem- 
ber of the population (the 155 rats as designated on 
p. 173) had with two groups. A sample of such 
an analysis is shown in figure 120. Six such 
analysis were made for the four ingroups and six 
for the random groups. These analyses provided 
data on 930 contacts with 2 ingroups, and 930 with 
2 random groups. These analyses are summarized 
in table 40. 
Messengers may be classified into three cate- 
gories on the basis of the number of contacts each 
has with two groups: 1. Few or no contacts with 
• ••••• ••• ••• •• 
•••••• •• •• 
I I I I I I I I iiii 
0 5 10 
NUMBER OF PAIRED ASSOCIATIONS 
WITH BOX 25 BURROW FEMALES 
Figure 120. — The number of paired associations each of the 
155 rats in the population had with members of two 
different ingroups. Each dot represents a rat. 
Table 40. — Proportion of population (755 rats) having 
contacts with 2 groups 
Frequency of contact 
With two 
ingroups 
With two 
random 
groups 
Contact with neither or only 1 
group 
0. 684 
0. 477 
Moderate (1 to 5) contacts with 
each group 
.263 
. 504 
Many contacts (6 or more) with 
1 group and moderate (1 or 
more) with the other 
.053 
. 018 
either group; 2. A moderate number of contacts 
with both groups; 3. Many contacts with one 
group and a moderate number with the other. 
Effectiveness as messengers should increase in the 
order listed. Organization of the population into 
ingroups increases the number of individuals who 
have little contact with either of two groups. 
In other words there is a decrease in the number 
of messengers. With regard to the latter two 
categories we will consider the effectiveness of 
rats as messengers. Effectiveness of transmission 
of a message should be dependent upon the 
frequency with which two rats make contact. 
This should be true whether we are considering 
m m 
> x 
co 
10 - 
o 
' -n 
: 2 
• £> 
m X) 
< m 
i-o 
> > 
O CO 
O CO 
m o 
X) O 
£> 
> IJ 
r~ o 
m 2 
CO CO 
5- 
OL 
175 
